- © 2002 by American Society of Clinical Oncology
Clinical Link Between p53and Angiogenesis in Lung Cancer
TWO TENETS OF modern tumor biology are that p53 alterations are among the most common genetic changes in human cancers and that induction of new blood vessels is required for tumors to exceed the limits of avascular growth and to metastasize.1,2 Therefore, it is not surprising that several laboratories have explored the intriguing hypothesis that p53 alterations would promote tumor neovascularization.3-5 An elegant genetic study of this in mice compared the growth and neovascularization of tumor xenografts from a p53 wild-type to a human colon adenocarcinoma cell line engineered to be null for p53.6 When compared with tumors from p53 wild-type cells, tumors produced by p53 null cells were palpable earlier, grew faster, and exhibited promotion of neovascularization. Establishing a clinical link between p53 status and angiogenesis in a consecutive series of resected non–small-cell lung carcinoma (NSCLC) cases is the subject of the article by Yuan et al7 in this issue of the Journal of Clinical Oncology. They explored this relationship in lung cancer because aberrant expression or mutations of p53 are frequent in invasive lung cancers and in bronchial preneoplasia.8,9 The analysis focused on the comprehensive evaluation of two angiogenic factors: vascular endothelial growth factor (VEGF) and interleukin 8 (IL-8). They report an association between aberrant p53 expression and neoangiogenesis, as assessed by tumor microvessel count and real-time quantitative reverse transcriptase polymerase chain reaction or immunohistochemical detection of VEGF and IL-8. Their work advances the field by demonstrating how aberrant p53 expression relates not only to a negative clinical outcome but also to high expression levels of these angiogenic factors.
The p53 tumor suppressor gene encodes a multifunctional transcription factor that is activated by stressful stimuli, including DNA damage and hypoxia.10 Genes that have a p53 response element include several involved in mediating cell cycle arrest or apoptosis, such as p21, BAX, GADD45, FAS, and insulin-like growth factor binding protein 3.10 Other p53 targets include potent regulators of angiogenesis, such as thrombospondin-1 and brain-specific angiogenesis inhibitor 1.3,11 It is proposed that in p53 null cells, the normal response to hypoxia is amplified through the stabilization of an oxygen-sensing transcription factor, hypoxia inducible factor 1-alpha subunit (HIF1-α).6 One of the targets of HIF1-α is VEGF. Binding of VEGF to receptors on vascular endothelial cells promotes proliferation and neovascularization.12 The chemokine IL-8 plays an important role in human tumor growth through mitogenic and angiogenic activities.13 Although a mechanistic link to p53 has yet to be established, IL-8 expression could be induced by stress present in the tumor environment, such as hypoxia or acidosis.
Neoangiogenesis occurs physiologically in wound healing and in adults in the reproductive organs of women. The fact that tumor progression is dependent on neoangiogenesis makes its inhibition an attractive therapeutic target. The ability to disrupt a tumor’s blood supply and inhibit its vascular growth could limit the potential for drug resistance or clinical toxicities observed with cancer chemotherapies. The considerable excitement generated by the angiogenesis inhibitors angiostatin and endostatin reflects the enthusiasm for this novel therapy.14 Several inhibitors of angiogenesis are in clinical trials and can be reviewed at the National Cancer Institute’s Cancer Trials Web site (http://cancertrials.nci.nih.gov/news/angio/). These include agents that block matrix breakdown (marimastat, BMS-275291), inhibit endothelial cells directly (endostatin, thalidomide), and antagonize stimulation of angiogenesis through the VEGF receptor (SU6668, anti-VEGF antibody).
VEGF expression is associated with a poor clinical outcome in breast cancer.15 The literature in NSCLC is less mature, but studies have assessed the prognostic impact of p53, VEGF, and microvessel count, with discordant results.7 Comparisons among these studies are complicated by differences in methodology and perhaps by tumor heterogeneity. For instance, the expression of angiogenic factors may vary based on tumor size and heterogeneous expression within a given tumor. Microvessel density could vary within tumors. Despite this, the tumor profiling reported by Yuan et al7 provides a strong scientific rationale to pursue clinical trials in lung cancer with antiangiogenic agents, perhaps emphasizing those that target VEGF activity.12
What is not known is how aberrant p53 expression or p53 mutations regulate VEGF or IL-8 expression. Whether this is a direct or indirect transcriptional effect needs to be determined. It will be interesting to learn how p53 coordinately regulates these angiogenic factors, which might engage common signaling pathways, as observed in vitro.16 If this occurs clinically, it would underscore a need for combination antiangiogenic therapy or the pharmacologic targeting of a common downstream pathway shared by both VEGF and IL-8.
The understanding of the association between p53, VEGF, and angiogenesis is relevant as inhibitors of angiogenesis enter clinical trials. The findings reported by Yuan et al7 set the stage for use of molecular profiling to select antiangiogenic therapies for lung cancer. Their work establishes a clinical link between p53 alterations and expression of specific angiogenic factors. Yet, many other genetic alterations occur in lung cancer that may promote angiogenesis, highlighting other targets for antiangiogenic therapy. Examples of this include alterations of the epidermal growth factor receptor (EGFR)/ras signaling pathway. Aberrant expression of the EGFR is frequent in NSCLC and in bronchial preneoplasia.9,17 In vitro response of anti-EGFR agents has been linked to antiangiogenic effects.18 Preliminary results of clinical trials with pharmacologic inhibitors of the EGFR-associated tyrosine kinase indicate antitumor activity in a variety of malignancies.19
In a sense, the work of Yuan et al leads us to an approach reminiscent of a criminal investigation where we build a profile (genetic alterations) and establish a modus operandi (neovascularization). We are entering an exciting phase in the molecular profiling of human tumors through the use of established prognostic markers, as well as the discovery of novel markers from microarray and proteomic analyses. Molecular profiling of cancers should provide biologic insights and attractive therapeutic targets based on pathways presumed essential for the growth of human tumors.
